Antibodies and superantibodies in patients with chronic rhinosinusitis with nasal polyps

Jiun-Bo Chen, PhD,a,b Louisa K. James, PhD,a,c Anna M. Davies, PhD,a,c Yu-Chang Bryan Wu, PhD,a,c Joanne Rimmer, FRCS,d Valerie J. Lund, MS, FRCS,d Jou-Han Chen, MS,b James M. McDonnell, PhD,a,c Yih-Chih Chan, PhD,a,c George H. Hutchins, BSci,a Tse Wen Chang, PhD,b Brian J. Sutton, DPhil,a,c Harsha H. Kariyawasam, FRCP, PhD,d and Hannah J. Gould, PhDa,c London, United Kingdom, and Taipei, Taiwan

Background: Chronic rhinosinusitis with nasal polyps is basophil degranulation, and IgG1 or -binding associated with local immunoglobulin hyperproduction and the fragments of each anti-SEE enhanced degranulation by the presence of IgE against Staphylococcus aureus other anti-SEE. enterotoxins (SAEs). Aspirin-exacerbated respiratory disease is Conclusions: SEEs can activate basophils by simultaneously a severe form of chronic rhinosinusitis with nasal polyps in binding as in the conventional manner to CDRs and as which nearly all patients express anti-SAEs. superantigens to framework regions of anti-SEE IgE in Objectives: We aimed to understand antibodies reactive to anti-SEE IgE-FcεRI complexes. Anti-SEE IgG1s can enhance SAEs and determine whether they recognize SAEs through the activity of anti-SEE IgEs as conventional antibodies through their complementarity-determining regions (CDRs) or CDRs or simultaneously as conventional antibodies and as framework regions. ‘‘superantibodies’’ through CDRs and framework regions to Methods: Labeled staphylococcal enterotoxin (SE) A, SED, and SEEs in SEE–anti-SEE IgE-FcεRI complexes. (J Allergy Clin SEE were used to isolate single SAE-specific B cells from the Immunol 2016;nnn:nnn-nnn.) nasal polyps of 3 patients with aspirin-exacerbated respiratory Key words: disease by using fluorescence-activated cell sorting. Chronic rhinosinusitis with nasal polyps, aspirin-exac- Recombinant antibodies with ‘‘matched’’ heavy and light chains erbated respiratory disease, Staphylococcus aureus enterotoxin, superantigen, superantibody, basophil were cloned as IgG1, and those of high affinity for specific SAEs, assayed by means of ELISA and surface plasmon resonance, were recloned as IgE and antigen-binding fragments. IgE activities were tested in basophil degranulation assays. Staphylococcus aureus and its superantigens are implicated in Results: Thirty-seven SAE-specific, IgG- or IgA-expressing the intense inflammatory processes of the upper and lower air- 1 B cells were isolated and yielded 6 anti-SAE clones, 2 each for ways in patients with allergic diseases. These superantigens, SEA, SED, and SEE. Competition binding assays revealed that in particular Staphylococcus aureus enterotoxins (SAEs), are the anti-SEE antibodies recognize nonoverlapping epitopes in strongly associated with chronic rhinosinusitis with nasal polyps SEE. Unexpectedly, each anti-SEE mediated SEE-induced (CRSwNP), particularly in the subpopulation of patients with aspirin-exacerbated respiratory disease (AERD), as well as those with , , and atopic dermatitis.2-5 SAEs are a family of structurally related proteins comprising different From athe Randall Division of Cell and Molecular Biophysics, King’s College London; serological types, such as staphylococcal enterotoxin (SE) A, bthe Genomics Research Center, , Taipei; cMRC & Asthma UK SEB, SEC, SED, and SEE (up to SEU) and toxic shock syndrome Centre for Allergic Mechanisms of Asthma, King’s College London, Guy’s Campus, toxin 1 (TSST-1).6 SAEs are potent T-cell superantigens, causing London; and dAllergy and Rhinology, Royal National Throat Nose Ear Hospital, London. polyclonal activation of up to 25% of certain T-cell populations Supported by the Medical Research Council (grant no. G1100090), the London Law by interacting with a common b-chain structural framework Trust (to L.K.J.), Medical Research Council (Y.-C.B. Wu), the National Institute for region in the T-cell receptor (TCR),7,8 rather than the Health Research (NIHR) Biomedical Research Centre at Guy’s and St Thomas’ complementarity-determining region (CDR), which recognizes NHS Foundation Trust and King’s College London (to H.J.G., B.J.S., and J.M.M.), and the Wellcome Trust for support of the King’s Biomolecular Spectroscopy Facility specific antigenic peptides bound to MHC. The activity of SEA (085944). The views expressed are those of the authors and not necessarily those of the and SED on B cells in vitro suggests that they can also act as National Health Service, the NIHR, or the Department of Health. B-cell superantigens by binding to the common structural Disclosure of potential conflict of interest: J.-B. Chen has received grants and support for framework regions in the immunoglobulin heavy-chain variable travel to meetings for the study of other purposes from the Genomics Research Center, region (VH) domains shared by immunoglobulins with Academia Sinica. V. J. Lund has received consultancy fees from GlaxoSmithKline, 9-11 Actelion, Vifor, Johnson & Johnson, and Navigent and payment for lectures from different CDRs, as previously shown for staphylococcal 12-14 Neilmed and MSD. The rest of the authors declare that they have no relevant conflicts protein A (SpA). This might increase the polyclonality of of interest. the B-cell repertoire in patients with CRSwNP and allow Received for publication September 25, 2015; revised May 7, 2016; accepted for publi- S aureus to escape immune surveillance.15,16 cation June 13, 2016. Corresponding author: Hannah J. Gould, PhD, Randall Division of Cell and Molecular Nasal polyps in patients with CRSwNP are inflammatory Biophysics, King’s College London, London SE1 1UL, United Kingdom. E-mail: outgrowths of the paranasal sinus mucosa, which are generally [email protected]. characterized by TH2 inflammation, local immunoglobulin 0091-6749 production, and eosinophil infiltration driven by IL-5 and Ó 2016 The Authors. Published by Elsevier Inc. on behalf of the American Academy of eotaxin.17-20 Up to 100% of patients with AERD express Allergy, Asthma & . This is an open access article under the CC BY li- cense (http://creativecommons.org/licenses/by/4.0/). anti-SAE IgEs in their nasal polyp homogenates and often have http://dx.doi.org/10.1016/j.jaci.2016.06.066 a higher prevalence of comorbid asthma and eosinophilic

1 2 CHEN ET AL J ALLERGY CLIN IMMUNOL nnn 2016

patients with CRSwNP, B cells expressing this IgE are confined Abbreviations used to the nasal mucosa, suggesting a role in sinonasal inflamma- 7-AAD: 7-Aminoactinomycin D tion.17 Whether the SAEs bind to CDRs, framework regions, or AERD: Aspirin-exacerbated respiratory disease both is addressed in the present work. CDR: Complementarity-determining region production in nasal polyps of patients with CRSwNP: Chronic rhinosinusitis with nasal polyps Fab: Antigen-binding fragment CRSwNP is driven by local activation and differentiation into FACS: Fluorescence-activated cell sorting plasma cells of B cells on exposure to various aeroallergens 26,27 SAE: Staphylococcus aureus enterotoxin and microbial antigens. Thus nasal polyps removed from SE: Staphylococcal enterotoxin patients with AERD provide a unique source of tissue to study SpA: Staphylococcal protein A how SAEs can shape the local antibody repertoire. In the first SPR: Surface plasmon resonance study of this kind, we used an efficient single-cell RT-PCR TCR: T-cell receptor method to clone and express antibodies from single SAE- TSST-1: Toxic shock syndrome toxin 1 specific B cells isolated from the nasal polyps of 3 patients VH: Heavy-chain variable region with AERD and tested their primary function in effector cell VL: Light-chain variable region activation. inflammation,17,21-23 and IgEs from nasal polyps activate 15 METHODS basophils in response to allergens and SEB in vitro. SEB acts Patients’ samples as a human T-cell superantigen in vivo to cause the symptoms 24 Nasal polyps and sera were collected from 3 patients with AERD of atopic dermatitis. Basophils isolated from patients with (HPK-014, HPK-016, and HPK-018) at the Royal National Throat, Nose atopic dermatitis with anti-SAE IgE in their sera, but not those and Ear Hospital, London, United Kingdom. The patients were male, with 25 from healthy control subjects, were responsive to SAEs. a mean age of 56 years. Only patient HPK-014 had positive skin prick Although anti-SAE IgE can be detected in the circulation of test responses to aeroallergens. The ethical committee representing

FIG 1. Isolation of SAE1 B cells from nasal polyps by means of FACS. Cells were labeled with a mixture of biotinylated SEA, SED, and SEE and phycoerythrin-coupled anti-CD138, allophycocyanin-coupled anti-CD19, and 7-AAD. A, The lymphocyte population was selected for size and granularity. FSC, Forward scatter; SSC, side scatter. B, Live B cells were identified as CD1917-AAD2 cells. C-E, Fluorescence minus one (FMO) control (Fig 1, C) and no SAE control (Fig 1, D) were used to select SAE1CD191CD13827- AAD2 cells, which constituted 0.25% of the total B-cell population (Fig 1, E). The No SAE control indicates that the cells were stained with all fluorescent antibodies and streptavidin–fluorescein isothiocyanate (FITC) but without biotinylated SAEs, whereas FMO control indicates cells stained with all antibodies but without streptavidin-FITC. J ALLERGY CLIN IMMUNOL CHEN ET AL 3 VOLUME nnn, NUMBER nn

FIG 2. Six of the expressed antibodies exhibited SAE binding. The cleared supernatants collected from 293T

cells, separately transfected with 40 IgG1 expression plasmids, were tested for reactivity with SEA, SEB, SEC1, SED, SEE, TSST-1, and ovalbumin (OVA) by means of ELISA. An antibody with specificity to a particular SAE was identified with an absolute OD value greater than that of the than control IgG. The antibodies selected for

further studies are shown in boldface. Control IgG is a human IgG1 with VH4 and Vk. We expected to select only high-affinity binders by using ELISA because of the low sensitivity of this assay. the Royal National Throat, Nose and Ear Hospital approved the study, SEC1, SED, SEE, and TSST-1 by using the UniCAP system (Phadia, and all patients provided written informed consent before the study Uppsala, Sweden; see Table E1 in this article’s Online Repository at commenced. www.jacionline.org).

Measurement of IgE to SAEs in sera and nasal polyp Single B-cell sorting by using fluorescence- homogenates activated cell sorting and single-cell RT-PCR of Detailed methods for processing sera and nasal polyp homogenates are immunoglobulin cDNA available in the Methods section in this article’s Online Repository at Single-cell suspensions from frozen nasal polyp samples were prepared by www.jacionline.org. All sera from blood samples and homogenates from 1 hour of digestion with 1 mg/mL hyaluronidase (Sigma-Aldrich, St Louis, nasal polyps were assayed for total IgE and specific IgE to SEA, SEB, Mo) and 1 U/mL Liberase TL (Roche, Mannheim, Germany) at 378C. Samples 4 CHEN ET AL J ALLERGY CLIN IMMUNOL nnn 2016

FIG 3. Specificity and high affinity of anti-SEE and anti-SEA antibodies. A-C, Recombinant IgG1 antibodies 1G2, 203, and 1F3 at indicated concentrations were tested for binding to SEA, SEB, SEC1, SED, SEE, and TSST-1 by using ELISA. D-F, Binding kinetics were characterized by means of SPR with biotin-labeled SEE or SEA immobilized on a streptavidin-coated sensor chip. Purified antibodies were injected at the indicated concentrations, followed by dissociation. were stained for CD19, CD138, SEA, SED, and SEE. SAE-specific single Sequence analysis of the monoclonal B cells were sorted by means of fluorescence-activated cell sorting (FACS) immunoglobulin genes into 96-well PCR plates for cDNA synthesis and immunoglobulin expression Immunoglobulin gene sequences were analyzed by using the IMGT/ cloning as IgG1 (Fig 1). Full details of the cloning and expression strategy are V-Quest tool,28 allowing identification of their clonal signatures (unique available in Fig E1 and Table E2 in this article’s Online Repository at CDR3 sequences); germline V, D, and J genes; and somatic mutations. www.jacionline.org. Sequences are available in GenBank (http://www.ncbi.nlm.nih.gov/ genbank), and their accession numbers are listed in Table E3 in this article’s Online Repository at www.jacionline.org. Antibody cloning, expression, and purification SEA-, SED-, and SEE-specific antibodies were expressed in human Surface plasmon resonance embryonic kidney 293T cells (ATCC, Manassas, Va) as IgG . Large-scale 1 Surface plasmon resonance (SPR) with a Biacore T200 instrument (GE production of IgE, IgG , and antigen-binding fragments (Fabs) was carried 1 Healthcare) was performed to determine binding specificity, kinetics, and out with Expi293 cells (Invitrogen, Carlsbad, Calif). The IgG antibodies 1 affinity of the antigen-antibody interactions. Further details are available in the were purified by means of protein A–Sepharose (GE Healthcare, Pittsburgh, Methods section in this article’s Online Repository. Pa), the IgE antibodies by -coupled NHS-activated Sepharose (GE Healthcare), and the Fabs by LambdaFabSelect agarose (GE Healthcare). Anti-SAE IgG1 antibodies were identified in culture supernatants by means of Basophil degranulation assay routine ELISA (Fig 2), and their purity was estimated by means of SDS-PAGE Degranulation of cells of the rat basophilic cell line RBL SX-38, which under reducing conditions (see Fig E2 in this article’s Online Repository at stably expresses the a, b, and g chains of human FcεRI,29 was used as a routine www.jacionline.org). Further details are available in the Methods section in method to measure the functionality of the SAE-specific antibodies. Further this article’s Online Repository. details are available in the Methods section in this article’s Online Repository. J ALLERGY CLIN IMMUNOL CHEN ET AL 5 VOLUME nnn, NUMBER nn

RESULTS Cloning of anti-SAE antibodies Three patients with AERD, HPK-014, HPK-016, and HPK-018, with titers of specific anti-SAEs in their serum or nasal polyp homogenate were studied (see Table E1). Specific IgEs to 6 common SAEs (SEA, SEB, SED, SEE, SEC, and TSST-1) were detected in all 3 homogenate samples and in the serum of patient HPK-018 and to 3 SAEs (SEA, SEE, and TSST-1) in the serum of patient HPK-14, whereas none were detected in the serum of patient HPK-016. Percentages of the anti-SAE IgEs were approximately 10-fold higher in the homogenates compared with the sera, ranging from 0.38% to 2.35% and 0.03% to 0.31%, respectively (see Table E1). Viable SAE-positive B cells (SAE1CD191CD1382 7-aminoactinomycin D [7-AAD]2) identified by means of FACS (Fig 1) were sorted into individual wells of 96-well plates. SAE-bound B cells accounted for 0.18%, 0.08%, and 0.25% of the nasal polyp B-cell population of patients HPK-014, HPK-016, and HPK-018, respectively (see Table E4 in this article’s Online Repository at www.jacionline.org). Matched immunoglobulin VH and light-chain variable region (VL; k or l) DNA pairs were amplified from the cDNAs of 37 SAE1CD191CD1382 single B cells (18 from patient HPK-014, 6 from patient HPK-016, and 13 from patient HPK-018; see Fig E1). Of these 37 VH sequences obtained, 23 were from g chains, and 14 were from a chains. No VH sequence was obtained from ε chains. The VH genes used in these 37 VH sequences consisted of 12 VH1 (32.43%), 13 VH3 (35.14%), 6 VH4 (16.22%), 5 VH5 (13.51%), and 1 VH6 (2.7%) sequences. Three (HPK-014_2D6, HPK- 018_2A11, and HPK-018_2C10) of the 37 B cells expressed both a k and a l chain. Of 40 VL sequences obtained, 25 were from the l chain, and 15 were from the k chain. All 37 B cells represented different clones ascertained by their unique CDR-H3 sequences (BioProject ID: PRJNA347912; https://www.ncbi.nlm.nih.gov/bioproject). We expressed all 40 of the putative SAE-specific matched variable regions as human IgG1 antibodies.

Identification and sequence analyses of anti-SAE antibodies All 40 antibodies, expressed as IgG1, were assayed for the 6 SAEs detected in nasal polyp homogenates by means of ELISA. Only SEA-, SED-, and SEE-specific antibodies could be identified, which is consistent with screening the B cells for only these specificities. Two anti-SEA antibodies, HPK-018_1B6 (IgA1, IGHV1-46) and HPK-016_1F3 (IgG1, IGHV1-69); 2 anti-SED antibodies, HPK-014_1A4 (IgA2, IGHV3-34) and HPK-018_2D6 (IgA1, IGHV3-49); and 2 anti-SEE antibodies, HPK-014_1G2 (IgA1, IGHV3-30) and FIG 4. Anti-SAE IgEs mediate SAE-induced basophil degranulation. A-C, HPK-016_203 (IgG4, IGHV5-51), were identified (Fig 2). The RBL SX-38 cells were sensitized with anti-SEA IgE 1B6 or 1F3 (Fig 4, A), germline immunoglobulin variable genes and isotypes used by anti-SED IgE 1A4 or 2D6 (Fig 4, B), or anti-SEE IgE 1G2 or 203 (Fig 4, C) those 6 antibodies are shown in Table E3. We refer to these and stimulated by the cognate SAE or anti-IgE as a positive control. D, RBL SX-38 cells were sensitized with the indicated anti-SAE IgEs and antibodies below as the anti-SEA antibodies 1B6 and 1F3, the stimulated with the indicated SAEs. Degranulation was assayed based on anti-SED antibodies 1A4 and 2D6, and the anti-SEE antibodies b-hexosaminidase release. 1G2 and 203. 6 CHEN ET AL J ALLERGY CLIN IMMUNOL nnn 2016

FIG 5. Anti-SEE antibodies bind to distinct nonoverlapping epitopes. A, 1G2 IgG1 (250 nmol/L) was injected onto a sensor chip coated with SEE for 3 minutes, immediately followed by a 3-minute injection of either

running buffer (black), 500 nmol/L 1G2 IgG1 (blue), or 100 nmol/L 203 IgG1 (red). B, After regeneration of the chip by using glycine-HCl (pH 2.5), the experiment was performed in reverse, with a 3-minute binding

of 100 nmol/L 203 IgG1, followed by a 3-minute injection of either buffer (black), 500 nmol/L 203 IgG1 (red), or 100 nmol/L 1G2 (blue).

The V(D)J rearrangement and CDR signatures of the 6 anti- induce cross-linking of FcεRI on the surfaces of basophils to SAE antibodies are shown in Table E3. A higher frequency of l cause basophil degranulation. The anti-SEAs (Fig 4, A) and the light chains (5/6) was found when compared with that in human anti-SEDs (Fig 4, B) behaved as expected, whereas both peripheral blood B cells (1/5).30 Sequence alignments of 6 anti-SEE IgEs, 1G2 and 203, were active in this assay (Fig 4, antibodies against their most closely related germline VH and C). Basophil activation would require either 2 identical epitopes VL gene sequences revealed a relatively high frequency of amino in the SEE because it formed a dimer like SED32 or 2 independent acid replacements in framework regions of VH and VL epitopes in an SEE molecule if it were a monomer. A recent chains (see Fig E3 in this article’s Online Repository at crystal structure of SEE in a complex with TCR revealed a mono- www.jacionline.org). meric structure.33 Thus we conclude that SEE has 2 distinct and nonoverlapping epitopes that explains this activity; this is illustrated and discussed below. SEC1 did not trigger 1G2 or Specificity and affinity of purified anti-SAE 203 IgE-mediated basophil degranulation, although both show antibodies cross-reactivity to SEC1 (Fig 4, D). ELISA was used to determine the specificity of the purified recombinant IgG1 antibodies with the SAEs (Fig 3, A-C, and results not shown). The 1G2 and 203 antibodies were specific Anti-SEE antibodies bind to nonoverlapping for SEE, although both displayed cross-reactivity with SEC1 epitopes on SEE (27% amino acid sequence homology with SEE) to a different To map the epitopes of anti-SEE 1G2 and 203 antibodies on extent rather than the more homologous SEA and SED (81% SEE, we carried out sandwich binding assays using ELISA and and 54% sequence homology with SEE, respectively).31 1F3 SPR. ELISA revealed that saturation of the 1G2 binding site, was specific for SEA but was not cross-reactive with SEE. All which was achieved through addition of SEE to immobilized 1G2 3 antibodies exhibited high affinity for their cognate SAEs, Fab and IgE, did not prevent binding of 203 IgG1 (see Fig E4 in demonstrating low nanomolar binding affinities, as measured this article’s Online Repository at www.jacionline.org). Also, from the kinetics of interaction with specific SAEs by using SPR revealed that binding of saturating concentrations of 1G2 SPR (Fig 3, D-F). IgG1 to immobilized SEE did not prevent the simultaneous binding of soluble 203 IgG1 (Fig 5, A). The reverse was also observed: after saturating all of the 203 binding sites on Anti-SEE IgE mediates SEE-induced basophil immobilized SEE, 1G2 IgG1 could still bind with high degranulation affinity (Fig 5, B). Both ELISA and SPR assays confirmed that The g1 chain constant region of the IgG1 antibodies was the 2 anti-SEE antibodies bind to nonoverlapping epitopes on substituted with the human ε constant region and the resulting SEE. IgEs were expressed and purified to examine the activities of the corresponding anti-SAE IgEs (see Fig E2, B). Basophil degranulation activities of the 6 anti-SAE IgEs in the presence Anti-SEE IgG1 modulates IgE-mediated of the cognate SAE are shown in Fig 4. degranulation Conventional antigen binding through CDRs to 2 identical The different anti-SEE isotypes, IgE and IgG1, were tested in receptor-bound monoclonal IgE molecules, each of which pairwise combinations in basophil degranulation assays to recognizes a single epitope in a monomeric antigen, should not characterize their interplay. As expected, because of competition J ALLERGY CLIN IMMUNOL CHEN ET AL 7 VOLUME nnn, NUMBER nn

inhibitory activity as the whole IgG1 antibodies (because of competitive inhibition) when Fab and IgE were of the same specificity. Cross-linking of adjacent IgE-SEE complexes on the basophil surface cannot occur with a monomeric Fab, even one that binds to the nonoverlapping epitope. Paradoxically, the Fab recognizing the nonoverlapping epitope retains the enhancing activity of the whole IgG1 (Fig 6, B). Fig 7, C, illustrates the mechanism by which receptor cross-linking can occur. We suggest that high-affinity binding of receptor-bound anti-SEE IgE to SEE through its CDRs to the epitope represented in green (Fig 7, C) facilitates subsequent low-affinity binding through a site on SEE, which is represented in red, that binds to a framework region of an adjacent IgG (or Fab) that recognizes a nonoverlapping SEE epitope, which is represented in blue. We call such an antibody, in principle of any isotype that acts synergistically to enhance IgE effector functions, a ‘‘superantibody.’’

DISCUSSION Several human mAbs against specific allergens have been obtained from allergic patients by means of phage display,36-40 EBV transformation of PBMCs,41 or single-cell RT-PCR.34 In addition, many groups have generated anti-SEB mAbs by using either mouse hybridomas or phage display and pursued their use as therapeutic agents.42,43 In earlier work we produced a recombinant allergen (Phl p 7)–specific antibody with the native (‘‘matched’’)heavy and light chain by using single-cell PCR.34 To the best of our knowledge, the present work is the first report of recombinant anti-SAE antibodies containing matched heavy and light chains cloned by using single-cell RT-PCR. We produced 6 anti-SAE antibodies that bind to SAEs with high affinity: 2B6 and 1F3 for SEA, 1A4 and 2D6 for SED, and 1G2 and 203 for SEE. The VH regions of 1F3 and 203 were derived from single IgG-expressing B cells and 1B6, 1A4, 1G2, and 2D6 from single IgA-expressing B cells, which were present in nasal polyps from 3 patients with CRSwNP (see Table E3). The antibodies were first expressed as IgG1 and then converted to IgE and Fab. We began with the hypothesis that SAEs can act as B-cell FIG 6. The unrelated anti-SEE IgG1 or Fab enhances anti-SEE IgE–mediated superantigens, as well as T-cell superantigens. Indeed, this basophil degranulation. RBL SX-38 cells were sensitized with anti-SEE or suggestion has often been made before. It is consistent with the anti-SEA IgE antibodies and stimulated with SEE in the presence of 1 1 1 evidence that SEA and SED bind weakly to immunoglobulins anti-SEE ( 1G2/ 203) or anti-SEA ( 1F3) IgG1 (A) or Fab (B). Dotted lines 10,11 represent positive (upper; anti-IgE) and negative (media only) controls. bearing VH3 and VH4, respectively. Several authors have Degranulation was assayed based on b-hexosaminidase release. demonstrated the overabundance of VH5 in IgE, suggesting that such a B-cell superantigen can recognize members of this VH 44-47 for epitope binding to SEE, the 1G2 IgG1 inhibited SEE-induced family. Low-affinity and promiscuous binding to the TCR degranulation of basophils sensitized with the 1G2 IgE, and the variable b (Vb) chains are characteristics of T-cell superantigens, 7,8 203 IgG1 inhibited degranulation of basophils sensitized with which include SEA, SED, and, notably, SEE. SEE binds to 6 34,35 48 the 203 IgE (Fig 6, A). As also expected, the IgG1 that different human Vb chains. We have shown that SEE can recognized the nonoverlapping epitope on IgE enhanced indeed behave as a B-cell superantigen by interacting with IgE degranulation (Fig 7, A) by cross-linking 2 SEE molecules bound molecules both conventionally through CDRs and unconvention- to adjacent receptor-bound IgE molecules; this is illustrated ally through framework regions (Fig 7, A). below in Fig 5, B. The 1F3 antibody, which is specific for SEA, ELISA assays used to select the 6 antibodies described in the had no such effects, demonstrating that antigen specificity is present study (Fig 3) would not have detected weak binding of crucial for the modulation of IgE activity by IgG1. SAEs to the framework regions. However, strong binding of SAEs to the CDRs in the context of a cell (Fig 2) would facilitate weak binding to the framework regions because of an increase in Paradoxical enhancing activity of anti-SEE Fabs avidity. Binding of SAEs to the framework regions as well as We repeated the pairwise assays by using the 2 anti-SEE IgG1 CDRs of BCRs can contribute to somatic hypermutation and Fabs (Fig 6, B). As expected, the Fabs exhibited the same affinity maturation of antibodies. Coker et al46 found 75% of 8 CHEN ET AL J ALLERGY CLIN IMMUNOL nnn 2016

mutational hot spots in the framework regions of IgE and IgA expressed by B cells in the nasal mucosa, which are often exposed to S aureus commensal infections.46 Superantigen-driven selection at the stage of local affinity maturation might explain the high frequency of mutations in the framework regions of our cloned anti-SAEs (see Fig E3). Due to the real-world limitations of the experimental methods used in this study, we have not obtained definitive evidence for the proposed interactions (Fig 7). First, IgE-expressing antibodies were not among the 40 recombinant antibodies we expressed; this was not surprising because of the extremely low abundance of IgE- compared with IgG- and IgA-expressing B cells.49 Second, although we have shown that IgG-expressing B cells undergo local switching to IgE in nasal polyps from patients with CRSwNP and in the respiratory tract mucosa in patients with allergic diseases,50-53 1G2 and 203 antibodies came from different patients. We would have needed IgE related to at least 1 of them and both antibody isotypes from the same subject to prove the existence of superantibodies in this AERD cohort. Nevertheless, in unpublished work using next-generation sequencing of B-cell repertoires in nasal polyps, we have observed as many as 17/10,000 IgE sequences related to IgG or IgA. In the B-cell repertoire of patient HPK-014, we were able to identify IgE relatives of 2 recombinant IgG1 antibodies (A: HPK-014_1A6 and B: HPK_014 2A8; see Fig E5 in this article’s Online Repository at www.jacionline.org), although these 2 IgG1 clones were not among the 6 clones that exhibited high affinity for the SAEs (Fig 3). One of the antibodies, 203, was originally an IgG4 and, interest- ingly, expresses IGHV5-51, the only of 2 VH5 family members that is overexpressed in IgE (Wu et al, unpublished results) and a candidate for binding in a superantigen-like manner.46 The other antibody, 1G2, was originally an IgA1 and expresses a member of the VH3 family (IGHV3-30, see Table E3). The CDRs differ between 203 and 1G2, and therefore it might be that SEE binds promiscuously to both VH5 and VH3 framework regions. Structural studies are required to define the epitopes in SEE. It was previously shown by using x-ray crystallography that SpA is recognized by a rheumatoid factor antibody in a superantigen-dependent manner, in which SpA interacts with the framework residues of the VH3 domain.14 We believe the anti- SEE superantibodies presented here might be the first antibodies shown to have both conventional antigen- and superantigen- binding sites and thus to have superantibody activity. The ability of the anti-SEE IgG1 and its Fabs to inhibit the activity of basophils sensitized by the related IgE resembles the blocking activities of IgG4 or IgA in specific allergen immunotherapy.34,35,54 Blocking IgG antibodies have been developed by many groups for therapeutic use as antitoxins. Nearly all antitoxin antibodies and all isolated anti-SEB anti- bodies are blocking, as opposed to enhancing, antibodies.42,43,55

FIG 7. Schematic representation of proposed antibody and ‘‘superanti- body’’ activities. A, A monomeric SEE molecule cross-links 2 FcεRI-bound = IgE molecules on the surface of a basophil (receptor not shown) through SEE molecules can affect this: one is recognized conventionally through 2 epitopes: the green site is recognized conventionally by the CDRs of the the 2 nonoverlapping (green and blue) epitopes by 2 unrelated anti-SEEs, IgE antibody, and the red site interacts with the framework regions of the and the other involves the superantigen site (red). Although the

IgE V region in a superantigen manner. B, An IgG1 antibody that recognizes cross-linking ability of the Fab reveals the interactions that underpin this a nonoverlapping blue epitope cross-links 2 FcεRI-bound IgE molecules that proposed mechanism, the whole antibody can act not only in the manner recognize the green epitope; both blue and green epitopes are recognized depicted in Fig 7, B, but also that shown in Fig 7, C. This antibody, exhibiting conventionally by the CDRs. C, Because the IgG1 Fab (highlighted in darker the ability to recognize 2 SEE molecules simultaneously through both CDRs blue) can also cross-link FcεRI-bound IgE, we propose that 2 monomeric and framework regions, we call a ‘‘superantibody.’’ J ALLERGY CLIN IMMUNOL CHEN ET AL 9 VOLUME nnn, NUMBER nn

SAE vaccination has been investigated, with encouraging success disease. However, similar antibodies can occur in patients with for efficacy in treating infections with S aureus.56,57 Whether allergic diseases and asthma associated with commensal S aureus anti-SEE IgGs from the present study might be similarly infections and dominant allergens. efficacious could be determined, but a combination of anti-SAEs might be even more efficacious than any single Clinical implications: Anti-SAE IgE antibodies alone or anti-SAE. together with IgG1 can contribute to the pathogenesis of To the best of our knowledge, 1G2 and 203 are the first CRSwNP and allergic disease. examples of enhancing anti-SAE antibodies. However, there is a precedent in the case of the antibody BAB2 against a dominant 58 birch pollen allergen, Bet v 1. BAB2 and, notably, its Fab REFERENCES enhanced binding of Bet v 1 to IgE and caused immediate-type 1. Huvenne W, Hellings PW, Bachert C. Role of staphylococcal superantigens in hypersensitivity reactions in human skin. The authors suggested airway disease. Int Arch Allergy Immunol 2013;161:304-14. that enhancing antibodies could account for the failure of specific 2. Rossi RE, Monasterolo G. Prevalence of serum IgE antibodies to the SAEs (SAE, 58-60 SEB, SEC, SED, TSST-1) in patients with persistent allergic rhinitis. Int Arch allergen immunotherapy. Although enhancing antibodies are Allergy Immunol 2004;133:261-6. relatively rare, they might be important because their activity 3. Van Zele T, Gevaert P, Watelet JB, Claeys G, Holtappels G, Claeys C, et al. might ‘‘win’’ in competition with an excess of blocking Staphylococcus aureus colonization and IgE antibody formation to enterotoxins antibodies. is increased in nasal polyposis. J Allergy Clin Immunol 2004;114:981-3. In previous studies it has been demonstrated that certain 4. Bachert C, van Steen K, Zhang N, Holtappels G, Cattaert T, Maus B, et al. Specific IgE against Staphylococcus aureus enterotoxins: an independent risk antibodies against a specific allergen are more frequently factor for asthma. J Allergy Clin Immunol 2012;130:376-81.e8. observed than others in atopic patients; such allergens, that is, 5. Leung AD, Schiltz AM, Hall CF, Liu AH. Severe atopic dermatitis is associated Der p 1 and Der p 2 in patients with house dust mite allergy, are with a high burden of environmental Staphylococcus aureus. Clin Exp Allergy said to be dominant allergens.61,62 These allergens usually have 2008;38:789-93. multiple epitopes, allowing more extensive cross-linking of the 6. Krakauer T, Stiles BG. The staphylococcal enterotoxin (SE) family: SEB and sib- ε lings. Virulence 2013;4:759-73. IgE-Fc RI complexes on effector cells. In this case noncompeti- 7. Herman A, Kappler JW, Marrack P, Pullen AM. Superantigens: mechanism of tive allergen-specific IgG or IgA, which are likely present in much T-cell stimulation and role in immune responses. Annu Rev Immunol 1991;9: higher concentrations than IgE, might be able to enhance cell 745-72. activation if the configuration of the 2 sites and the resulting 8. Thomas D, Chou S, Dauwalder O, Lina G. Diversity in Staphylococcus aureus enterotoxins. Chem Immunol Allergy 2007;93:24-41. distance between cross-linked complexes is favorable (Fig 7, B). 9. Levinson AI, Kozlowski L, Zheng Y, Wheatley L. B-cell superantigens: definition 63 SEE was first documented in a case of food poisoning. It was and potential impact on the immune response. J Clin Immunol 1995;15(suppl): later identified as a 26.4-kDa single-chain protein belonging to the 26S-36S. group III superantigens (the SEA superfamily), having strong 10. Domiati-Saad R, Attrep JF, Brezinschek HP, Cherrie AH, Karp DR, Lipsky PE. homology to SEA and SED, as stated above. SEE has been shown Staphylococcal enterotoxin D functions as a human B cell superantigen by 64,65 rescuing VH4-expressing B cells from apoptosis. J Immunol 1996;156:3608-20. to be a potent enterotoxin and a polyclonal activator of T cells 11. Domiati-Saad R, Lipsky PE. Staphylococcal enterotoxin A induces survival of 48 and binds promiscuously to multiple human TCR Vb chains. VH3-expressing human B cells by binding to the VH region with low affinity. Anti-SEE IgE was demonstrated in 16.7% of nasal polyps J Immunol 1998;161:1257-66. compared with 0% of control tissue, as well as in all nasal polyp 12. Kristiansen SV, Pascual V, Lipsky PE. Staphylococcal protein A induces biased 17 production of Ig by VH3-expressing B lymphocytes. J Immunol 1994;153: samples from all 3 patients with CRSwNP in the present study. 2974-82. Sequence comparison of SEA, SEC1, SED, and SEE and 13. Kozlowski LM, Kunning SR, Zheng Y, Wheatley LM, Levinson AI. Staphylo- inspection of the SEA and SEE crystal structures32,33 suggest coccus aureus Cowan I-induced human immunoglobulin responses: preferential that certain surface-exposed residues could alter the electrostatic IgM rheumatoid factor production and VH3 mRNA expression by protein A- binding B cells. J Clin Immunol 1995;15:145-51. potential or shape of the enterotoxin surface, rendering SEE and 14. Graille M, Stura EA, Corper AL, Sutton BJ, Taussig MJ, Charbonnier JB, et al. SEC1 different from SEA and SED and thereby contributing to Crystal structure of a Staphylococcus aureus protein A domain complexed with the cross-reactivity of the 203 antibody. the Fab fragment of a human IgM antibody: structural basis for recognition of Basophils were activated to release granular mediators by B-cell receptors and superantigen activity. Proc Natl Acad Sci U S A 2000;97: allergens and SEB when sensitized by IgE in homogenates from 5399-404. 15 15. Zhang N, Holtappels G, Gevaert P, Patou J, Dhaliwal B, Gould H, et al. Mucosal the nasal polyps of patients with CRSwNP in vitro, indicating tissue polyclonal IgE is functional in response to allergen and SEB. Allergy 2011; the functionality of IgE antibodies in nasal polyps. Thus the 66:141-8. putative anti-SEE IgEs 1G2 and 203 in patients HPK_014 and 16. Bachert C, Zhang N. Chronic rhinosinusitis and asthma: novel understanding of the role of IgE ‘above atopy’. J Intern Med 2012;272:133-43. HPK_016, together with the potential IgG4 or IgA1 precursors, 17. Gevaert P, Holtappels G, Johansson SG, Cuvelier C, Cauwenberge P, Bachert C. could have activated a variety of effector cells and contribute to Organization of secondary lymphoid tissue and local IgE formation to Staphylo- the proinflammatory environment in the patients’ nasal polyps. coccus aureus enterotoxins in nasal polyp tissue. Allergy 2005;60:71-9. This is compatible with evidence from the work of Gevaert 18. Van Zele T, Gevaert P, Holtappels G, van Cauwenberge P, Bachert C. Local et al66 showing that omalizumab is therapeutic in patients with immunoglobulin production in nasal polyposis is modulated by superantigens. CRSwNP, although an earlier study did not show efficacy.67 It Clin Exp Allergy 2007;37:1840-7. 19. Hulse KE, Norton JE, Suh L, Zhong Q, Mahdavinia M, Simon P, et al. Chronic should also be noted that IgG and IgA superantibody activities rhinosinusitis with nasal polyps is characterized by B-cell inflammation and EBV- are independent of isotypes. induced protein 2 expression. J Allergy Clin Immunol 2013;131:1075-83, e1-7. In summary, our results provide proof of concept that SAES 20. Kato A. Immunopathology of chronic rhinosinusitis. Allergol Int 2015;64:121-30. can act as B-cell superantigens and that anti-SAE antibodies of 21. Bachert C, Gevaert P, Holtappels G, Johansson SG, van Cauwenberge P. Total and specific IgE in nasal polyps is related to local eosinophilic inflammation. any isotype can behave as superantibodies. We have isolated J Allergy Clin Immunol 2001;107:607-14. antibodies with potential superantibody activity from nasal 22. Bachert C, Zhang N, Holtappels G, De Lobel L, van Cauwenberge P, Liu S, et al. polyps of patients with AERD, a highly inflammatory TH2 Presence of IL-5 protein and IgE antibodies to staphylococcal enterotoxins in 10 CHEN ET AL J ALLERGY CLIN IMMUNOL nnn 2016

nasal polyps is associated with comorbid asthma. J Allergy Clin Immunol 2010; 45. Snow RE, Djukanovic R, Stevenson FK. Analysis of VH tran- 126:962-8, e1-6. scripts in a bronchial biopsy of an asthmatic patient confirms bias towards VH5, 23. Tomassen P, Jarvis D, Newson R, Van Ree R, Forsberg B, Howarth P, et al. Staph- and indicates local clonal expansion, somatic mutation and isotype switch events. ylococcus aureus enterotoxin-specific IgE is associated with asthma in the general Immunology 1999;98:646-51. population: a GA(2)LEN study. Allergy 2013;68:1289-97. 46. Coker HA, Harries HE, Banfield GK, Carr VA, Durham SR, Chevretton E, et al. 24. Skov L, Olsen JV, Giorno R, Schlievert PM, Baadsgaard O, Leung DY. Applica- Biased use of VH5 IgE-positive B cells in the nasal mucosa in allergic rhinitis. tion of staphylococcal enterotoxin B on normal and atopic skin induces up- J Allergy Clin Immunol 2005;116:445-52. regulation of T cells by a superantigen-mediated mechanism. J Allergy Clin Im- 47. Wu YC, James LK, Vander Heiden JA, Uduman M, Durham SR, Kleinstein SH, munol 2000;105:820-6. et al. Influence of seasonal exposure to grass pollen on local and peripheral blood 25. Leung DY, Harbeck R, Bina P, Reiser RF, Yang E, Norris DA, et al. Presence of IgE repertoires in patients with allergic rhinitis. J Allergy Clin Immunol 2014; IgE antibodies to staphylococcal exotoxins on the skin of patients with atopic 134:604-12. dermatitis. Evidence for a new group of allergens. J Clin Invest 1993;92:1374-80. 48. Kumar S, Menoret A, Ngoi SM, Vella AT. The systemic and pulmonary immune 26. Kato A, Hulse KE, Tan BK, Schleimer RP. B-lymphocyte lineage cells and the response to staphylococcal enterotoxins. Toxins (Basel) 2010;2:1898-912. respiratory system. J Allergy Clin Immunol 2013;131:933-58. 49. Dullaers M, De Bruyne R, Ramadani F, Gould HJ, Gevaert P, Lambrecht BN. The 27. De Schryver E, Devuyst L, Derycke L, Dullaers M, Van Zele T, Bachert C, et al. who, where, and when of IgE in allergic airway disease. J Allergy Clin Immunol Local immunoglobulin e in the nasal mucosa: clinical implications. Allergy 2012;129:635-45. Asthma Immunol Res 2015;7:321-31. 50. Coker HA, Durham SR, Gould HJ. Local somatic hypermutation and class switch 28. Brochet X, Lefranc MP, Giudicelli V. IMGT/V-QUEST: the highly customized recombination in the nasal mucosa of allergic rhinitis patients. J Immunol 2003; and integrated system for IG and TR standardized V-J and V-D-J sequence anal- 171:5602-10. ysis. Nucleic Acids Res 2008;36:W503-8. 51. Takhar P, Smurthwaite L, Coker HA, Fear DJ, Banfield GK, Carr VA, et al. 29. Wiegand TW, Williams PB, Dreskin SC, Jouvin MH, Kinet JP, Tasset D. High- Allergen drives class switching to IgE in the nasal mucosa in allergic rhinitis. affinity oligonucleotide ligands to human IgE inhibit binding to Fc epsilon recep- J Immunol 2005;174:5024-32. tor I. J Immunol 1996;157:221-30. 52. Takhar P, Corrigan CJ, Smurthwaite L, O’Connor BJ, Durham SR, Lee TH, et al. 30. van der Burg M, Tumkaya T, Boerma M, de Bruin-Versteeg S, Langerak AW, van Class switch recombination to IgE in the bronchial mucosa of atopic and nona- Dongen JJ. Ordered recombination of immunoglobulin light-chain genes topic patients with asthma. J Allergy Clin Immunol 2007;119:213-8. occurs at the IGK locus but seems less strict at the IGL locus. Blood 2001;97: 53. Gevaert P, Nouri-Aria KT, Wu H, Harper CE, Takhar P, Fear DJ, et al. Local re- 1001-8. ceptor revision and class switching to IgE in chronic rhinosinusitis with nasal 31. Balaban N, Rasooly A. Staphylococcal enterotoxins. Int J Food Microbiol 2000; polyps. Allergy 2013;68:55-63. 61:1-10. 54. James LK, Shamji MH, Walker SM, Wilson DR, Wachholz PA, Francis JN, et al. 32. Rodstrom KE, Regenthal P, Lindkvist-Petersson K. Structure of staphylococcal Long-term tolerance after allergen immunotherapy is accompanied by selective enterotoxin E in complex with TCR defines the role of TCR loop positioning persistence of blocking antibodies. J Allergy Clin Immunol 2011;127:509-16, in superantigen recognition. PLoS One 2015;10:e0131988. e1-5. 33. Sundstrom M, Abrahmsen L, Antonsson P, Mehindate K, Mourad W, Dohlsten M. 55. Dutta K, Varshney AK, Franklin MC, Goger M, Wang X, Fries BC. Mechanisms The crystal structure of staphylococcal enterotoxin type D reveals Zn21- mediating enhanced neutralization efficacy of staphylococcal enterotoxin B by mediated homodimerization. EMBO J 1996;15:6832-40. combinations of monoclonal antibodies. J Biol Chem 2015;290:6715-30. 34. James LK, Bowen H, Calvert RA, Dodev TS, Shamji MH, Beavil AJ, et al. 56. Schlievert PM. The dream of staphylococcal vaccination. J Exp Med 2014;211: Allergen specificity of IgG(4)-expressing B cells in patients with grass 2326. pollen allergy undergoing immunotherapy. J Allergy Clin Immunol 2012;130: 57. Spaulding AR, Salgado-Pabon W, Merriman JA, Stach CS, Ji Y, Gillman AN, 663-70.e3. et al. Vaccination against Staphylococcus aureus pneumonia. J Infect Dis 2014; 35. Dodev TS, Bowen H, Shamji MH, Bax HJ, Beavil AJ, McDonnell JM, et al. In- 209:1955-62. hibition of allergen-dependent IgE activity by antibodies of the same specificity 58. Eibensteiner P, Denepoux S, Steinberger P, Kraft D, Visco V, Banchereau J, et al. but different class. Allergy 2015;70:720-4. Expression of a human IgG4 antibody, BAB2, with specificity for the major Birch 36. de Lalla C, Tamborini E, Longhi R, Tresoldi E, Manoni M, Siccardi AG, et al. pollen allergen, Bet v 1 in Escherichia coli: recombinant BAB2 Fabs enhance the Human recombinant antibody fragments specific for a rye-grass pollen allergen: allergic reaction. Int Arch Allergy Immunol 1999;118:190-2. characterization and potential applications. Mol Immunol 1996;33:1049-58. 59. Denepoux S, Eibensteiner PB, Steinberger P, Vrtala S, Visco V, Weyer A, et al. 37. Steinberger P, Kraft D, Valenta R. Construction of a combinatorial IgE library Molecular characterization of human IgG monoclonal antibodies specific for from an allergic patient. Isolation and characterization of human IgE Fabs with the major birch pollen allergen Bet v 1. Anti-allergen IgG can enhance the specificity for the major timothy grass pollen allergen, Phl p 5. J Biol Chem anaphylactic reaction. FEBS Lett 2000;465:39-46. 1996;271:10967-72. 60. Selb R, Eckl-Dorna J, Vrtala S, Valenta R, Niederberger V. An assay that may 38. Flicker S, Steinberger P, Norderhaug L, Sperr WR, Majlesi Y, Valent P, et al. Con- predict the development of IgG enhancing allergen-specific IgE binding during version of grass pollen allergen-specific human IgE into a protective IgG(1) anti- birch immunotherapy. Allergy 2013;68:1199-202. body. Eur J Immunol 2002;32:2156-62. 61. Cui Y. Immunoglobulin E-binding epitopes of mite allergens: from characteriza- 39. Persson H, Sadegh MK, Greiff L, Ohlin M. Delineating the specificity of an IgE- tion to immunotherapy. Clin Rev Allergy Immunol 2014;47:344-53. encoding transcriptome. J Allergy Clin Immunol 2007;120:1186-92. 62. Thomas WR. Hierarchy and molecular properties of house dust mite allergens. 40. Levin M, Davies AM, Liljekvist M, Carlsson F, Gould HJ, Sutton BJ, et al. Hu- Allergol Int 2015;64:304-11. man IgE against the major allergen Bet v 1—defining an epitope with limited 63. Morris CA, Conway HD, Everall PH. Food-poisoning due to staphylococcal cross-reactivity between different PR-10 family proteins. Clin Exp Allergy enterotoxin E. Lancet 1972;2:1375-6. 2014;44:288-99. 64. Merkenschlager M, Fisher AG. Selective manipulation of the human T-cell recep- 41. Visco V, Dolecek C, Denepoux S, Le Mao J, Guret C, Rousset F, et al. Human tor repertoire expressed by thymocytes in organ culture. Proc Natl Acad Sci U S IgG monoclonal antibodies that modulate the binding of specific IgE to birch pol- A 1992;89:4255-9. len Bet v 1. J Immunol 1996;157:956-62. 65. Yan XJ, Li XY, Imanishi K, Kumazawa Y, Uchiyama T. Study of activation of 42. Chow SK, Casadevall A. Monoclonal antibodies and toxins—a perspective on murine T cells with bacterial superantigens. In vitro induction of enhanced re- function and isotype. Toxins (Basel) 2012;4:430-54. sponses in CD41 T cells and of anergy in CD81 T cells. J Immunol 1993; 43. Varshney AK, Wang X, Scharff MD, MacIntyre J, Zollner RS, Kovalenko OV, 150:3873-81. et al. Staphylococcal Enterotoxin B-specific monoclonal antibody 20B1 success- 66. Gevaert P, Calus L, Van Zele T, Blomme K, De Ruyck N, Bauters W, et al. Oma- fully treats diverse Staphylococcus aureus infections. J Infect Dis 2013;208: lizumab is effective in allergic and nonallergic patients with nasal polyps and 2058-66. asthma. J Allergy Clin Immunol 2013;131:110-6.e1. 44. Snow RE, Chapman CJ, Frew AJ, Holgate ST, Stevenson FK. Pattern of usage 67. Pinto JM, Mehta N, DiTineo M, Wang J, Baroody FM, Naclerio RM. A random- and somatic hypermutation in the V(H)5 gene segments of a patient with asthma: ized, double-blind, placebo-controlled trial of anti-IgE for chronic rhinosinusitis. implications for IgE. Eur J Immunol 1997;27:162-70. Rhinology 2010;48:318-24. J ALLERGY CLIN IMMUNOL CHEN ET AL 10.e1 VOLUME nnn, NUMBER nn

METHODS recombination with a Geneart Seamless Assembly Kit (Invitrogen) according Preparation of sera and nasal polyp homogenates to vendor’s instructions as shown in Fig E1. Briefly, 20 ng of paired VH and VL DNA segments, 50 ng of backbone Blood samples were allowed to clot at room temperature for 1 hour and insert, and 100 ng of linear vector were combined with Seamless enzyme mix centrifuged at 2000g for 20 minutes, and serum was separated and stored at and incubated at room temperature for 30 minutes. The assembly products 2708C. PBS containing protease inhibitor cocktail (Roche) was added to were immediately transformed into competent E coli (One Shot Top 10, Invi- snap-frozen nasal polyp specimens at 1 mL per 0.1 g of tissue. Tissues trogen), and transformants were analyzed for the presence of DNA inserts by were homogenized with a TissueLyser (Qiagen, Hilden, Germany) for using colony PCR. The plasmid DNA of positive clones was purified and 2 minutes at 30 Hz until disrupted, homogenates were centrifuged at sequenced to ensure no base mutations and correct recombination. 15,000g for 20 minutes at 48C, and supernatants were collected and stored Human embryonic kidney 293T cells (ATCC) were seeded in 6-well plates at 2708C. to produce cloned antibodies on a small scale for validating their binding specificities. Transient transfections were performed at 80% cell confluency Single cell sorting by using FACS by linear polyethylenimine with an average molecular weight of 25 kDa (Polysciences, Warrington, Pa) as a transfection reagent. The cells were After dispersion by means of enzymatic digestion of nasal polyps, cells cultured for 2 days before the supernatants were harvested and analyzed by were washed with complete RPMI medium (10% FBS in RPMI medium) to means of ELISA for SAE binding reactivity. The concentrations of recom- remove the enzymes, filtered through a 40-mm cell strainer, and finally binant IgG antibodies released into the supernatants of transfected 293T cells resuspended at a concentration of 1 3 106 cells/mL in FACS buffer (5% 1 were measured by means of ELISAwith human IgG Quantitation Kits (Bethyl normal goat serum [Invitrogen, Carlsbad, Calif] and 2 mmol/L EDTA in Laboratories, Montgomery, Tex). The g1 constant regions of SAE-specific PBS). Then cells were stained with phycoerythrin-coupled anti-CD138 antibody expression vectors were replaced by the ε constant region or the (BioLegend, San Diego, Calif), allophycocyanin-coupled anti-CD19 (BD C 1 domain of the g1 constant region, to construct IgE and Fab expression Biosciences, San Jose, Calif), 7-AAD (eBioscience, San Diego, Calif), and H vectors, respectively. biotinylated SAEs (Toxin Technology, Sarasota, Fla) in FACS buffer. Biotinylated SEA, SED, and SEE were simultaneously used for labeling cells each at a concentration of 10 mg/mL and detected by means of FACS with ELISA for SAE-specific antibodies in culture streptavidin–Alexa Fluor 488 (Invitrogen) at a dilution of 1:1000. supernatants Microtiter plates were coated with 2 mg/mL of each SEA, SEB, SEC1, SED, SEE, and TSST-1 (Toxin Technology) or ovalbumin (50 mL/well) Single-cell RT-PCR overnight at 48C and blocked with 200 mL of assay diluent (0.5% BSA, 0.05% Single SAE1CD191CD13827-AAD2 B cells were directly sorted into Tween-20, and 0.01% thimerosal in PBS). Bound IgG1 antibodies were de- 96-well PCR plates (Fermentas, Vilnius, Lithuania) containing 18 mL/well tected by using horseradish peroxidase–conjugated goat anti-human IgG Fc 3 of ice-cold 1 RT buffer (Invitrogen) containing 0.1% Triton-X 100, antibody (Jackson ImmunoResearch, West Grove, Pa) and developed with 10 mmol/L dithiothreitol, 300 ng of random hexamers (Invitrogen), NeA-Blue TMB substrate (Clinical Science Products, Mansfield, Mass). 0.5 mmol/L dNTP mix (Fermentas), and 40 U of Ribolock RNase Inhibitor (Fermentas) on a FACSAria II cell sorter (BD Biosciences). Fifty units of Superscript III reverse transcriptase (Invitrogen) was added, and then reverse Basophil degranulation assay 8 8 8 transcription was performed at 42 C for 10 minutes, 25 C for 10 minutes, 50 C RBL SX-38 cells were typically grown in complete MEM medium, as 8 for 60 minutes, and 85 C for 5 minutes. previously described.E3 One day before a planned experiment, RBL Primer nucleotide sequences to cover all known V-region genes and SX-38 cells were seeded at 6 3 105 cells/mL (500 mL/well) in 48-well plates. reverse primers specific for k and l constant regions were as previously After an overnight incubation, 250 mL of complete MEM medium containing E1 ε described. Reverse primers for g, a, and constant regions were Cg1 SAE-specific IgE at 1 mg/mL was added to sensitize the cells. After a 2-hour 9 9 ε 9 (5 -GGAAGGTGTGCACGCCGCTGGTC-3 ), C 1(5-GGTTTTGTTGTC incubation at 378C, each well was washed twice with 500 mL of prewarmed 9 9 GACCCAGTCTG-3 ), and Ca1(5-TGGGAAGTTTCTGGCGGTCACG- Tyrode buffer (135 mmol/L NaCl, 5 mmol/L KCl, 5.6 mmol/L 9 9 3 ) as primary primers for the first-round PCR and Cg2(5-GTTCGGGG glucose, 1.8 mmol/L CaCl , 1 mmol/L MgCl , 20 mmol/L HEPES, and 9 9 2 2 AAGTAGTCCTTGAC-3 ), Ca2(5-GTCCGCTTTCGCTCCAGGTCACA 0.5 mg/mL BSA, pH 7.3) and then 250 mL of Tyrode buffer containing 9 ε 9 9 CT-3 ), and C 2(5-TGGCATAGTGACCAGAGAGCGTG-3 ) as nested SAEs at 1 mg/mL or polyclonal anti-human IgE antibody (Bethyl Labora- primers for the second-round PCR, respectively. All PCRs were carried out tories) at 3 mg/mL or 1% Triton X-100 was added to the cells. After a by proofreading Phusion DNA polymerase (Fermentas). Second-round PCR 30-minute incubation in a 378C incubator, culture media were collected and products were sequenced after purification by using the QIAquick PCR subjected to centrifugation at 300g for 5 minutes at room temperature. Fifty Purification Kit (Qiagen). microliters of cleared supernatants were transferred to each well of a 96-well black OptiPlate (PerkinElmer, Waltham, Mass), and then 50 mLof Recombinant antibody cloning and expression substrate solution (80 mmol/L 4-methylumbelliferyl-N-acetyl-D-glucosami- Purified second-round PCR products were used as a PCR template for nide [Sigma-Aldrich] in 0.1 mol/L citric acid buffer, pH 4.5) was added to amplifying DNA inserts by V-region– and J-region–specific primers contain- each well. The tape-sealed plate was incubated at 378C for 1 hour, and the re- ing an overhang sequence (Table E2), which are homologous to the bases at action was terminated by adding 100 mL of glycine buffer (0.2 mol/L glycine E2 and 0.2 mol/L NaCl, pH 10.7). The resulting fluorescence (excitation, 355 nm; one end of adjacent DNA fragments of pIgG1(k) or pIgG1(l) vectors. Purified DraIII/NheI- and BspEI/NheI-digested DNA fragments from emission, 460 nm) was measured with a Victor 3 fluorescence reader (PerkinElmer). Measured values were expressed as percentages relative to pIgG1(k) and pIgG1(l) vectors were used as a template for amplifying ‘‘back- bone insert’’ by CL forward primers, SL-Ck1 or SL-Cl1, and the IgG leader the value of release (100%) obtained by lysing cells with 1% Triton X-100. reverse primer SL-IgG Leader- (Fig E1 and Table E2), respectively. All PCR products above were synthesized by using Phusion DNA polymerase and pu- SPR rified with the QIAquick PCR Purification Kit. The linear pIgG1(k) and Purified and concentrated immunoglobulin solutions were quantified with a pIgG1(l) vectors were prepared by means of double digestion with DraIII/ NanoDrop spectrophotometer (Thermo Scientific, Wilmington, Del) by using NheI and BspEI/NheI, respectively, and purified by means of agarose gel elec- absorbance at 280 nm, and the protein concentration was calculated with the trophoresis and gel extraction (gel purification kit, Qiagen). The assembly of extinction coefficient of each antibody, which was predicted by conducting the antibody expression vector was carried out by means of homologous computation with a ProtParam program (http://web.expasy.org/protparam/). 10.e2 CHEN ET AL J ALLERGY CLIN IMMUNOL nnn 2016

Biotin-labeled SEE and SEA were immobilized on a streptavidin-coated REFERENCES sensor chip (GE Healthcare). Binding of SAE-specific IgG1 antibodies was E1. James LK, Bowen H, Calvert RA, Dodev TS, Shamji MH, Beavil AJ, et al. measured at 258C by using a 3-minute association phase and a 10-minute Allergen specificity of IgG(4)-expressing B cells in patients with grass pollen dissociation phase (unless otherwise specified in the figure legend) in a allergy undergoing immunotherapy. J Allergy Clin Immunol 2012;130: 663-70.e3. concentration series ranging from 250 to 16 nmol/L. One antibody was E2. Chen JB, Wu PC, Hung AF, Chu CY, Tsai TF, Yu HM, et al. Unique epitopes on injected for 3 minutes, followed by a 3-minute injection of the second antibody C epsilon mX in IgE-B cell receptors are potentially applicable for targeting or buffer, to detect simultaneous binding of 2 antibodies. After regeneration IgE-committed B cells. J Immunol 2010;184:1748-56. with glycine-HCl (pH 2.5), the experiment was performed in reverse. Standard E3. Shiung YY, Chiang CY, Chen JB, Wu PC, Hung AF, Lu DC, et al. An anti-IgE double-referencing data subtraction methods were used with a control surface monoclonal antibody that binds to IgE on CD23 but not on high-affinity IgE.Fc (biotin alone). receptors. Immunobiology 2012;217:676-83. J ALLERGY CLIN IMMUNOL CHEN ET AL 10.e3 VOLUME nnn, NUMBER nn

FIG E1. Cloning strategy for constructing recombinant IgG1 expression vector. Paired VH and VL regions from sorted single SAE-specific B cells were amplified by means of RT-PCR and sequenced as described in the Methods section in this article’s Online Repository. The 15-bp homologous arm was

introduced by means of PCR to permit cloning into pIgG1(k) or pIgG1(l) vectors by means of homologous recombination with the Seamless method, as described in the Methods section in this article’s Online

Repository. CL, Constant region of the light chain; JH, J region of the heavy chain; JL, J region of the light chain; L, leader; SL, seamless. 10.e4 CHEN ET AL J ALLERGY CLIN IMMUNOL nnn 2016

FIG E2. Purity of recombinant SAE-specific IgG1, IgE, and Fab antibodies. All recombinant SAE-specific IgG1 (A), IgE (B), and Fab (C) antibodies were transiently expressed in Expi293F cells and purified by means of protein A, anti-IgE, or LambdaFabSelect affinity chromatography, respectively. Five micrograms of each antibody was fractionated by using 12% SDS-PAGE under reducing conditions before Coomassie Blue staining. J ALLERGY CLIN IMMUNOL CHEN ET AL 10.e5 VOLUME nnn, NUMBER nn

FIG E3. Amino acid sequence alignment for 6 cloned SAE-specific antibodies against germline sequences. The V genes encoding the VH (A) and VL (B) of SAE-specific antibodies were identified by using IMGT/V-Quest. Somatic hypermutations that resulted in amino acid changes are indicated. 10.e6 CHEN ET AL J ALLERGY CLIN IMMUNOL nnn 2016

FIG E4. 1G2 and 203 bind to distinct epitopes on SEE. Each well of a 96-well ELISA plate was coated with 1G2 Fab or IgE antibodies at 1 mg/mL (50 mL/well), followed by saturating 1G2 antibody binding sites with SEE at

1 mg/mL (100 mL/well). The 1F3, 1G2, and 203 IgG1 antibodies were added to each well, followed by incubation with peroxidase-conjugated goat anti-human IgG/Fc. J ALLERGY CLIN IMMUNOL CHEN ET AL 10.e7 VOLUME nnn, NUMBER nn

FIG E5. Expressed single-cell IgG1 clones with IgE relatives identified by using next-generation sequencing (NGS) of B-cell repertoire. The immunoglobulin VH gene repertoire in the nasal polyps from patient HPK-014 was captured on the Illumina 2x300 MiSeq System and bioinformatically analyzed (manuscript

in preparation). The sequences in the HPK-014 NGS data set were aligned with the recombinant IgG1 antibodies expressed from 18 FACS-sorted SAE1 B cells (SAE1CD191CD13827-AAD2), and their clonal relatedness was determined by using clone-specific CDR3 DNA motifs. Two clonal families were identified,

showing 2 expressed recombinant IgG1 antibodies (A: HPK-014_1A6 and B: HPK-014_2A8; Sequence ID in blue) related to IgE (sequence ID in red) and other IgG mutants in the HPK-014 NGS data set. For illustration purposes, not all members of the clones isolated by using NGS are shown. 10.e8 HNE AL ET CHEN

TABLE E1. SAE-specific IgEs in serum and nasal polyp homogenate Serum Nasal polyp homogenate Patient ID Total IgE SEA SEB SED SEE SEC TSST-1Total IgE SEA SEB SED SEE SEC TSST-1 HPK-014 89.3 0.17* (0.19%) — — 0.28 (0.31%) — 0.16 (0.18%) 26.1 0.25 (0.96%) 0.12 (0.46%) 0.1 (0.38%) 0.13 (0.5%) 0.13 (0.5%) 0.21 (0.8%) HPK-016 53.2 — — — — — — 8.1 0.11 (1.36%) 0.11 (1.36%) 0.08 (0.99%) 0.09 (1.11%) 0.11 (1.36%) 0.19 (2.35%) HPK-018 317 0.09 (0.03%) 0.81 (0.26%) 0.78 (0.25%) 0.13 (0.04%) 0.05 (0.02%) 0.07 (0.02%) 48 0.59 (1.23%) 0.24 (0.5%) 0.31 (0.65%) 0.27 (0.56%) 0.32 (0.67%) 0.28 (0.58%)

*Kilounits per liter was introduced to express the level of IgE. Percentage of total IgE was the percentage of each SAE-specific IgE in total IgE. LEG LNIMMUNOL CLIN ALLERGY J nnn 2016 J ALLERGY CLIN IMMUNOL CHEN ET AL 10.e9 VOLUME nnn, NUMBER nn

TABLE E2. Primers used for antibody cloning Primer Sequence

VH cloning PCR SL-VH1/3/5f gttgctacgcgtgtcCTGAGCSAGGTGCAGCTGGTGSAGTC SL-VH1-3f gttgctacgcgtgtcCTGAGCCAGGTCCAGCTTGTGCAGTC SL-VH1-18f gttgctacgcgtgtcCTGAGCCAGGTTCAGCTGGTGCAGTC SL-VH1-24f gttgctacgcgtgtcCTGAGCCAGGTCCAGCTGGTACAGTCTG SL-VH2f gttgctacgcgtgtcCTGAGCCAGGTCACCTTGARGGAGTCTG SL-VH3-23f gttgctacgcgtgtcCTGAGCGAGGTGCAGCTGTTGGAGTCT SL-VH4f gttgctacgcgtgtcCTGAGCCAGSTGCAGCTGCAGGAGT SL-VH4-34f gttgctacgcgtgtcCTGAGCCAGGTGCAGCTACARCAGTGG SL-VH6f gttgctacgcgtgtcCTGAGCCAGGTACAGCTGCAGCAGTCA SL-VH7f gttgctacgcgtgtcCTGAGCCAGGTGCAGCTGGTGCAAT Vk cloning PCR SL-Vk1f ggctcccaggtgcacgatgtgACATCCAGWTGACCCAGTCTCC SL-Vk2f ggctcccaggtgcacgatgtgATATTGTGATGACCCAGACTCCACTCT SL-Vk3f ggctcccaggtgcacgatgtgAAATTGTGTTGACRCAGTCTCCAG SL-Vk4f ggctcccaggtgcacgatgtgACATCGTGATGACCCAGTCTC SL-Vk5f ggctcccaggtgcacgatgtgAAACGACACTCACGCAGTCTC SL-Vk6f ggctcccaggtgcacgatgtgAAATTGTGCTGACTCAGTCTCCA Vl cloning PCR SL-Vl1-40/50/51f tccttgcttatgggtccggaGTGGATTCTCAGTCTGTGYTGACGCAGCC SL-Vl1-36/44/47f tccttgcttatgggtccggaGTGGATTCTCAGTCTGTGCTGACTCAGCCA SL-Vl2f tccttgcttatgggtccggaGTGGATTCTCAGTCTGCCCTGACTCAGCC SL-Vl3f tccttgcttatgggtccggaGTGGATTCTTCCTATGAGCTGACWCAGCCA SL-Vl3-19f tccttgcttatgggtccggaGTGGATTCTTCTTCTGAGCTGACTCAGGAC SL-Vl4/5/9f tccttgcttatgggtccggaGTGGATTCTCAGSCTGTGCTGACTCAGCC SL-Vl4-60/69f tccttgcttatgggtccggaGTGGATTCTCAGYCTGTGCTGACTCAATC SL-Vl6f tccttgcttatgggtccggaGTGGATTCTAATTTTATGCTGACTCAGCCC SL-Vl7/8f tccttgcttatgggtccggaGTGGATTCTCAGRCTGTGGTGACYCAGGAG SL-Vl10f tccttgcttatgggtccggaGTGGATTCTCAGGCAGGGCTGACTCAGCC JH cloning PCR SL-JH1/4/5r gatgggcccttggtgctagcTGAGGAGACGGTGACCAGG SL-JH2r gatgggcccttggtgctagcTGAGGAGACAGTGACCAGGGT SL-JH3r gatgggcccttggtgctagcTGAAGAGACGGTGACCATTGTC SL-JH6r gatgggcccttggtgctagcTGAGGAGACGGTGACCGTG Jk cloning PCR SL-Jk1r tgcagccaccgtacgTTTGATTTCCACCTTGGTCCCT SL-Jk2r tgcagccaccgtacgTTTGATCTCCAGCTTGGTCCCT SL-Jk3r tgcagccaccgtacgTTTGATATCCACTTTGGTCCCA SL-Jk4r tgcagccaccgtacgTTTGATCTCCACCTTGGTCCCT SL-Jk5r tgcagccaccgtacgTTTAATCTCCAGTCGTGTCCCTT Jl cloning PCR SL-Jl1r ggccttgggctgacctaggACGGTGACCTTGGTCC SL-Jl2/3r ggccttgggctgacctaggACGGTCAGCTTGGTCC SL-Jl6r ggccttgggctgacctaggACGGTCACCTTGGTGC SL-Jl7r ggccttgggctgacctaggaCGGTCAGCTGGGTGC Backbone insert cloning PCR

SL-Ck1 CGTACGGTGGCTGCACCATC SL-Cl1 GGTCAGCCCAAGGCCAACC SL-IgG Leader- GACACGCGTAGCAACAGCGA

Homologous sequences to the bases at one end of the adjacent DNA fragment are displayed in lower case. 10.e10 CHEN ET AL J ALLERGY CLIN IMMUNOL nnn 2016

TABLE E3. Germline immunoglobulin variable gene and CDR3 sequences of cloned VH and VL pairs in the SAE-specific antibodies Accession Accession Clone name Isotype (VH) IGHV IGHD IGHJ CDR-H3 amino acids (VL) IGK(L)V IGK(L)J CDR-L3 amino acids HPK-014_1A4 A (a2) KT369789 3-23 3-9 3 CAKKRRYDAVTGYLFDMW KT369790 4-1 (k) 2 CQQFYSPAPYTF HPK-014_1G2 A (a1) KT369791 3-30 6-13 4 CARDRAQGIVAAAGTSFGYW KT369792 3-25 (l) 2 CQSGDTSGFDPDVAF HPK-016_203 G (g4) KT369793 5-51 2-15 4 CVRHCSGGRCSSAPLDSW KT369794 6-57 (l) 3 CLSHDTTNFVF HPK-018_1B6 A (a1) KT369795 1-46 2-8 3 CAYWGADGFDIW KT369796 1-44 (l) 3 CAVWDDSLSAWVF HPK-018_1F3 G (g1) KT369797 1-69 3-9 4 CARGMTGYVGAYDYW KT369798 1-44 (l) 3 CASWDDRLNGPVF HPK-018_2D6 A (a1) KT369799 3-49 2-2 4 CTTYCSSTSCQIDYW KT369800 2-14 (l) 1 CSSYTGSSGYVV

The VH and VL DNA sequences of each clone are available in GenBank (http://www.ncbi.nlm.nih.gov/genbank), and accession numbers are indicated in the table. IGHV, Immunoglobulin heavy-chain variable region; IGKV, immunoglobulin kappa-chain variable region; IGLV, immunoglobulin lambda-chain variable region. J ALLERGY CLIN IMMUNOL CHEN ET AL 10.e11 VOLUME nnn, NUMBER nn

TABLE E4. Cell populations of nasal polyp tissues HPK-014 HPK-016 HPK-018 Lymphocytes* 10.1% (255,337) 7.42% (157,133) 17.2% (313,277) Live cells (7-AAD2) 94.3% (239,878) 87.1% (136,067) 95.3% (297,159) Live B cells (CD1917-AAD2) 20.7% (52,717) 21.1% (33,024) 29.4% (91,607) SAE1 cells (SAE1CD1382)à 0.18% (92) 0.08% (26) 0.25% (225) Plasma cells (SAE2CD1381)à 2.98% (1,573) 0.1% (34) 0.37% (339)

*164% of total nasal polyp cells. 165% of lymphocytes. à166% of total live B cells.